Withdrawal of life support in the neurological intensive care unit

Objective: To describe the frequency and clinical course of terminal extubation in the neurological intensive care unit, to identify factors that influence the decision to withdraw life support, and to evaluate the experiences of surrogate decision-makers. Background: The right of patients to refuse life-prolonging treatment is widely accepted. However, it is unknown how frequently critically ill neurologic patients are removed from life support, and practice guidelines for withdrawing mechanical ventilation remain poorly defined. Methods: We reviewed the medical records of all patients cared for by the Columbia-Presbyterian neurocritical care service over a 3-year period who died, and identified a subgroup of non-brain-dead patients who were terminally extubated. We retrospectively analyzed the clinical course of these patients and interviewed their surrogate decision-makers. Results: Of 74 non-brain-dead patients, 32 (43%) were terminally extubated. Hispanic and white patients were more likely to be extubated than were African American patients (p = 0.02). The median duration of survival after extubation was 7.5 hours; 25% died within 1 hour, and 69% within 24 hours. Depth of coma did not predict the duration of survival after extubation. The most frequent signs after extubation were agonal or labored breathing (59%) and tachypnea (34%). Morphine or fentanyl was given to relieve respiratory distress in 68% of cases; the average dose of morphine was 6.3 mg/hour (range 2.5 to 20 mg/hour). In a structured interview of 24 surrogate decision-makers, 88% were satisfied or very satisfied with the overall process, and 75% felt the patient suffered minimally before death; all but one (96%) said that they would repeat the decision to withdraw life support. Conclusions: Forty-three percent of our non-brain-dead patients who died were terminally extubated. The duration of survival after extubation exceeded 24 hours in one third, and was not predicted by level of consciousness. Two thirds of patients were treated with opioids for agonal respiratory distress. Most surrogate decision-makers were comfortable and satisfied with the process of withdrawing care.